Relationship of instruments and policy in the Hawaii warning system

Abstract

The tsunami which follows a strong local earthquake occurs within a few minutes of the origin time. This requires that any warning system for such local tsunamis be decentralized, as in the system used by the Japan Meterological Agency. Thus, decisions are being made by local officials rather than in a central office. This usually means that the level of training, the education, and the preparation is not as intense in such a local office as in a central office. Therefore, the decision making should be assisted by instrumentation that discriminates tsunamigenic earthquakes from non‐tsunamigenic earthquakes. This is not yet possible in real time; however, since only very large earthquakes (more than 6.5) generate significant tsunamis, an instrument to inform the local official of the tsunami prospects can be implemented.

An instrument for assisting the local decision maker has been developed. This instrument consists of an analog computer (an inverted pendulum having a period of 0.75 sec and damping about 0.3) and a digital computer (a hardwired signal‐recognition circuit), providing output to a display of status or alarm. The level of displacement (or velocity or acceleration) , the number of times that the threshold must be exceeded, and the time window within which the excedance must occur are all adjustable. Initial settings require about 0.06 g three times, not more than ten seconds apart. Battery back‐up and test circuitry are provided.

Since the operating instructions require that motion be felt before the instrument alarm be considered valid, great weight is given to assuring that the instrument will operate when required. The instru‐ment is always “ON,”; with the earthquake turning it “OFF”; thus, it is continually self‐testing.

Twelve units have been installed in police stations, fire stations, or similar locations around the State of Hawaii, which funded the system developed at Indiana University. This tsunami seismic trigger should be considered whenever the primary objective is to trigger an alarm rather than to record data; the emphasis in design and development has been on reminding the local official when there may be a tsunami hazard and not on recording research data.